Those look like nice NOS sockets. They will be good as long as you can make a good solder joint on the opposite side. In fact, I think a gap is better for heat management.
Started testing my build.....
I am using a different PT and OT than what Pete originally perscirped. Bigger PT and OTs for more power.
I am using 1R cathode resistors rather than the 10R Pete used. As such his 400mV cathode voltage should be 40mV in my build.
My B+ is 450V DC fully loaded with all tubes in. Heaters read a nice 6.3V AC.
I use the bias trim pots and set -35V on the 6JN6s grid stoppers.
When I go to read the voltage across the 1R 6JN6 cathode resistors I am reading ~0.8V on all tubes.
Not sure what I am doing wrong?
I am using a different PT and OT than what Pete originally perscirped. Bigger PT and OTs for more power.
I am using 1R cathode resistors rather than the 10R Pete used. As such his 400mV cathode voltage should be 40mV in my build.
My B+ is 450V DC fully loaded with all tubes in. Heaters read a nice 6.3V AC.
I use the bias trim pots and set -35V on the 6JN6s grid stoppers.
When I go to read the voltage across the 1R 6JN6 cathode resistors I am reading ~0.8V on all tubes.
Not sure what I am doing wrong?
So turns out the wonky bias readings were due to my DMM ground being connected directly to chassis ground rather than audio ground on the PCB. Those little diodes and their Vf threw things off.
But now the biasing of the 6JN6 is giving me troubles. The trim pot jumps from 30mV to 60mV, I can't get it to settle in the 35-55mV range where I want the bias to be (aiming for 40mV).
With a B+ of 450V would the 6JN6 bias current still be 40mV?
But now the biasing of the 6JN6 is giving me troubles. The trim pot jumps from 30mV to 60mV, I can't get it to settle in the 35-55mV range where I want the bias to be (aiming for 40mV).
With a B+ of 450V would the 6JN6 bias current still be 40mV?
At 450 V on the anode you are dissipating 18 watts. It includes screen dissipation, so it is probably not exceeding the 17.5 watt limit on plate dissipation. What is you screen voltage and actual screen dissipation? I would think you would be better off in the long run at say 35 ma per tube. I hope Pete or George would tell us.
Can’t think what is causing the instability in you bias measured across the 1 Ohm resistor. Have you excercised the pots a bit?
Can’t think what is causing the instability in you bias measured across the 1 Ohm resistor. Have you excercised the pots a bit?
I looked around for a 10k multi-turn pot, but can't find any with similar pin spacing compared to stock.
jeff
Thanks for the helpful advice everyone!
I tried moving around the bias trim pots this morning and I think it helped. I had never thought of exercising trim pots, learned a new one. The left channel I was now able to stabilize at ~40mV!
The right channel however, no luck. After ~15mV I can move the bias trim in the smallest tiny increment but it still jumps to 50+ mV and then the tubes start going in to thermal run away brining the B+ down with it.
So right now I am a bit stuck. I don't know if I have a weak set of 6JN6 on the right hand side or the trim pot is just too sensitive.
I only have four 6JN6, so my next idea is to bring the known working "good" 6JN6 from the left hand channel to the right and see if I can bias those up there.
My only theory on why biasing is tricky is I did not use the stock PT. In my case the PT I have has a 130V - 0V bias winding (no CT). As such, I am taking that 130V - 0V winding to a bridge rectifier then straight in to a 1K dropping resistor (should I have added a cap on this node!?!). This setup gives me -60V the same as what Pete perscribes so I dont think this should be an issue. The only thing is maybe that extra 1K of resistance is messing with the bias string (?)
I tried moving around the bias trim pots this morning and I think it helped. I had never thought of exercising trim pots, learned a new one
. The left channel I was now able to stabilize at ~40mV! The right channel however, no luck. After ~15mV I can move the bias trim in the smallest tiny increment but it still jumps to 50+ mV and then the tubes start going in to thermal run away brining the B+ down with it.
So right now I am a bit stuck. I don't know if I have a weak set of 6JN6 on the right hand side or the trim pot is just too sensitive.
I only have four 6JN6, so my next idea is to bring the known working "good" 6JN6 from the left hand channel to the right and see if I can bias those up there.
My only theory on why biasing is tricky is I did not use the stock PT. In my case the PT I have has a 130V - 0V bias winding (no CT). As such, I am taking that 130V - 0V winding to a bridge rectifier then straight in to a 1K dropping resistor (should I have added a cap on this node!?!). This setup gives me -60V the same as what Pete perscribes so I dont think this should be an issue. The only thing is maybe that extra 1K of resistance is messing with the bias string (?)
A snappy jump in bias current can be caused by several things. A wonky pot is one, but a sharp jump from 15 mA to 50 mA can be oscillation. If you have the GNFB connected, disconnect it, at least until you get things working. A scope can see oscillation if you have one, but I often used a portable AM radio tuned to a weak station sitting near the board. If the radio goes bonkers when you turn the bias pot, you have oscillation. If moving your hands around near the tubes and transformer wires (be careful not to touch anything conductive) causes a change in current, you have oscillation. Try rerouting the transformer wires if oscillation is suspected.
After everything is working I would set the idle current to a point just above where the crossover distortion just becomes inaudible or the THD at say half a watt levels off. I was always squeezing these boards for every last watt, so I used 30 mA with 6HJ5's on 600+ volts. Once everything is working, start around 30 ma and listen for a while, then increase to 35, listen, then go up to 40. Go back to the last current where no improvement was heard. Decent quality sweep tubes can live quite a while at their maximum dissipation rating. They will live longer at a lower current. It's a trade off between lifetime and sound quality.
ok I think I am making progress lol
I tried swapping 6JN6 tubes from one channel to the other, the bias issue remains on the right channel even after a output tube swap. So I am going to say the problem not a tube issue.
George as per your suggestion I disconnected the GNFB on the right channel. Amp seems to bias up no problem now.
So I have an oscillation issue I guess?
Where do I go from here? I can post pics of my build but the construction is clean and as per Pete's build.
Could I have issue with the preamp tubes? Worth replacing those?
I tried swapping 6JN6 tubes from one channel to the other, the bias issue remains on the right channel even after a output tube swap. So I am going to say the problem not a tube issue.
George as per your suggestion I disconnected the GNFB on the right channel. Amp seems to bias up no problem now.
So I have an oscillation issue I guess?
Where do I go from here? I can post pics of my build but the construction is clean and as per Pete's build.
Could I have issue with the preamp tubes? Worth replacing those?
If the OPT is wired out of phase it is possible for GNFB to be GPFB, IE positive feedback cresting an oscillator. When bringing up a new amp I always minimize the possibilitise for error and feedback of any kind is a possible error source. With this board I fired it up as built without GNFB and it worked. I built three boards total. The first one was built as close to Petes plans as I could with the parts that I had on hand. I used it as my test mule and general science lab for over 3 years (post #138 in 2009 to post #1431 in 2014). It and one of the other two were packed in a box when I moved out of Florida in 2014. They have not seen power since then. Whenever I tried something new I first brought it up without GNFB, then tried it. The board as designed is marginal on gain at power levels in the 50+ WPC range. GNFB reduces the gain further. Using a 6GU5 for the driver tube buys some more gain. Pete switched to the 6GU5 in his 50 Watt mono block version.
During that time it existed in several flavors from 15 WPC to 250 WPC. The other two were built for 125 WPC and stayed that way. All of them sounded best without GNFB to my ears on my speakers. YMMV.
During that time it existed in several flavors from 15 WPC to 250 WPC. The other two were built for 125 WPC and stayed that way. All of them sounded best without GNFB to my ears on my speakers. YMMV.
Ok I will try keeping the GNFB connected on the PCB but while flipping the leads of the right side OT, hopefully it is just GPFB that needs to turn to GNFB by flipping of the OT primaries.
It would be weird though that the right side OT is backwards no?
Both OTs are the same, have the same wire colour coding and correspond to the wire colours marked on the PCB.
It would be weird though that the right side OT is backwards no?
Both OTs are the same, have the same wire colour coding and correspond to the wire colours marked on the PCB.
Ok it was GPFB that was causing the ossicllation/bias soaring issues on the right channel!
I moved the NFB wire coming off the PCB from the red output binding post to the black output binding post and all was good on the right channel. I will get to further burn in and testing later but at least now the bias is stable and not flying off in to thermal run down.
Is there any difference in:
1. Having the NFB wire coming off the PCB on the black output binding post
vs.
2. Haing the NFB wire coming off the PCB on the red output binding post and swapping the OT primary leads
???
I moved the NFB wire coming off the PCB from the red output binding post to the black output binding post and all was good on the right channel. I will get to further burn in and testing later but at least now the bias is stable and not flying off in to thermal run down.
Is there any difference in:
1. Having the NFB wire coming off the PCB on the black output binding post
vs.
2. Haing the NFB wire coming off the PCB on the red output binding post and swapping the OT primary leads
???
Triple posting, ignore my previous posts!!
The amp is up and running, found I had GPFB on both channels. So for now I just removed the feedback and I am running the amp with no feedback until I have time to swap the OT primary leads.
The amp is up and running, found I had GPFB on both channels. So for now I just removed the feedback and I am running the amp with no feedback until I have time to swap the OT primary leads.
1. How important are the preamp voltages? Due to my higher B+ I have adjusted R48 and R49 from 1K to 16K. This puts the preamp plate voltages at ~160V, still a bit high from the specified 130V in the schematic. http://www.pmillett.com/file_downloads/dcpp_sch.pdf
2. How important is it that the bias voltage C- is -60V for the preamp? My PT bias winding is higher than what is speced in the schematic. As such I have had to drop the bias voltage with a dropping resistor. I will make sure this dropping resistor is specified appropriately if it is critical that C- is -60V to get the preamp tubes CCS to operate optimally.
3. What heat sink have people used on this build? Placing a Mouser order for some other odds and ends and want to buy a suitable heatsink for the chassis mounted TO-220 devices.
You can just swap the secondary leads on the OPTs if you want.
Voltages aren't terribly important as long as the tube is within it's power envelope.
Buying a heatsink from Mouser will be EXPENSIVE... If you're in the GTA, I'd suggest checking out A1 Parts... http://www.a1parts.ca
If they work using the chassis as a sink, why change it?
Voltages aren't terribly important as long as the tube is within it's power envelope.
Buying a heatsink from Mouser will be EXPENSIVE... If you're in the GTA, I'd suggest checking out A1 Parts... http://www.a1parts.ca
If they work using the chassis as a sink, why change it?
Good suggestion, I love A1! I started a new job that brings me close to it often, it's great